Seberg & Linde-Laursen Systematic Botany 21:3-15 (1996)
Eremium was erected in 1994 to accommodate what had hitherto been known as Elymus erianthus. In arguing for recognition of the genus. Seberg and Linde-Laursen devoted most of the discussion to showing why the species should be excluded from Elymus. They looked at morphological, ecological, cytological (C- and N-banding), and molecular characteristics (rbcL sequence variation). Eremium erianthus differs from other species of Elymus in its long ciliate lemmas. It is unusual, but by no means unique, in having linear-subulate glumes. The authors note that such glumes are found in Leymus but state that its inclusion in that genus "would only serve to make Leymus more heterogenous".
The species is endemic to Argentina and grows in in dry to very dry steppe or in scree.
The cytological studies showed the species to be a hexaploid (2n=42). C-bands were distributed similarly on all chromosomes and occured primarily around the centromeres. In this the E. erianthus genome differs from those of the S and N genomes, both of which have predominantly telomeric and distal bands. The C-banding patterns were less conclusive, but suggested that the species is an alloploid rather than an autoploid.
The rbcL data were somewhat equivocal. The semistrict concensus tree clearly excluded Eremium from Elymus, but did not suggest where it belonged.